Abstract: The overall goal of project 3 of the University of Minnesota (UMN) Udall Center is to investigate why deep brain stimulation (DBS) therapy for Parkinson's disease (PD) works better in some individuals than in others and to develop methods to decrease the variability of DBS therapy for individual motor signs of PD. To deliver DBS therapy at a level consistent with the best responders, it is critical to investigate the (1) emergence of DBS- induced side effects that impede the delivery of more effective stimulation parameters, (2) logistical challenges in optimizing stimulation settings for each parkinsonian motor sign on an individual basis, and (3) multi-scale neurophysiological differences across the basal ganglia, thalamus, and brainstem that underlie the individual variability to DBS therapy. This project will leverage the well-characterized non-human primate model of PD (systemic MPTP) implanted with two scaled-down versions of the human DBS lead (subthalamic nucleus, STN and globus pallidus, GP). The approach involves a novel combination of high-field imaging (7T/10.5T, Imaging Core), computational neuron modeling of DBS, development of optimization algorithms based on quantitative behavioral assessments, multi-parameter regression analysis techniques (Biostatistics Core), and multi-scale electrophysiological analysis of DBS therapy that spans single-cell, ensemble, and whole-brain levels. Aim 1 will investigate the ability for narrow DBS pulse widths to extend the therapeutic parameter space window between alleviating parkinsonian motor signs and evoking motor side-effects. This aim will further enhance our understanding of the functional relationships between DBS parameter settings and their resultant therapeutic effect sizes and wash-in/wash-out time constants on a subject-specific, pathway-specific basis. Aim 2 will develop a novel real-time, behavior-based optimization algorithm for automatic and efficient selection of DBS parameters that minimize the expression of individual parkinsonian motor signs including rigidity, bradykinesia, akinesia, and gait/posture. Aim 3 will identify the subject-specific electrophysiological features that most closely correlate with the temporal and steady-state behavioral responses to DBS found in the first two aims. The simultaneous recordings will include local field potentials in the STN and GP as well as unit-spike recordings in three nuclei innervated by pallidofugal projection neurons (i.e. motor thalamus, centromedian-parafascicular complex of thalamus, and pedunculopontine nucleus). At the conclusion of the experiments, whole-brain transcription factor analysis for two metabolic markers (c-fos and egr-1) will be conducted through histological techniques to provide single-cell resolution for the neural pathways modulated by behaviorally-optimized DBS therapy. Together, these aims will provide critical new insight into the pathophysiological basis for the expression of each parkinsonian motor sign and which specific targeted pathways and electrophysiological features are most relevant to delivering the most effective and efficient level DBS therapy for each individual.